Financial Services Bot
This case study details the successful implementation of a foundational, rules-based chatbot that solved critical business needs. It laid the groundwork for future innovation. See the next-generation ** a simulated version ** Gen AI version here.
Website and Mobile Chatbot
Project type
Rules-based chatbot for a Financial Services firm
Team, Role and Responsibility
Team: 1 PM, 2 Data Scientists, 4 Engineers, 1 Scrum Master
My Role: Lead Conversation Designer
Tools Used
🧩 Problem
The firm’s contact center was overwhelmed by a high volume of customer inquiries related to keeping up with bills, checking finances, and fixing something urgent. This led to long wait times, poor customer experiences, and lost trust. The existing support systems couldn’t scale fast enough to meet the growing demand.
Q: What were the constraints?
-
Tight implementation window. We needed to reduce call volume within 60–90 days post-implementation.
-
Legacy systems. New features in the Financial Services Bot had to integrate with existing APIs and Dialogflow ES, limiting the complexity of responses and context handling. Migration in Dialogflow CX was in progress.
-
Security and compliance. Financial institutions use cases require clear audit trails, limited data storage, and strict fallback handling.
What Customers Were Trying to Do
-
Keep up with bills:
"I lost my debit card." "I saw a charge I don't recognize." -
Check finances:
"How much did I spend this month?" "What's my balance?" -
Fix something urgent:
"I lost my debit card." "I saw a charge I don't recognize."
Business Impact:
-
Rising call abandonment rates
-
Negative client sentiment in satisfaction surveys
-
High agent burnout and operational inefficiencies
💡Solution
I designed a solution that achieved 34% fewer handoffs, 23% more containment, 18% faster resolutions, and a 4.2/5 user rating post-launch.
New Features
-
Modified user account settings
-
Tracked and summarized spending habits
-
Sent bill reminders and processed payments
-
Locked misplaced debit cards
-
Answered FAQs and routed complex queries to live agents
📊 Outcomes
-
📉 34% reduction in live agent handoffs (within 90 days post-launch)
-
📈 23% increase in containment rate (self-service without escalation)
-
⏱️ 18% faster resolution for common financial tasks (e.g., balance inquiries, payments)
-
😀 4.2/5 average user satisfaction rating (based on in-chat thumbs up/down prompts)
🛠 My Contributions
-
Led end-to-end design from discovery through deployment
-
Created flow maps and dialog logic in Lucidchart and Figma
-
Developed sample scripts and utterance libraries for Dialogflow ES and CX
-
Co-developed escalation triggers and fallback messages with Engineering
🧠 My Mental Model
Q: Why did I approach it this way?
-
I focused on narrow yet high-frequency use cases to drive ROI
-
I used a modular design approach, so flows could scale or evolve as business needs change.
-
Partnered closely with engineering to prototype and test quickly.
Q: What did I learn from the solution?
-
Trust is the anchor of relationships in financial interactions. Tone and transparency are equally important as accuracy.
-
Modular flows and clear re-entry points improve flexibility and self-service.
-
Post-launch feedback loops (e.g., in-bot feedback) are critical for continuous fine tuning and personalization.
🎯 Impact
Financial Services Bot relieved pressure on the contact center while improving service consistency across digital channels. In addition to cost savings, and speed, it laid the groundwork for future AI-focused service expansions,
